Systems in data centers or companies have been integrated by cloud migration recently. As described in FIG. 1, a cloud system includes virtualized servers (or storages) and network switches connecting them (hereinafter, abbreviated to switches), and is often connected to a core network (a router, a L3 (Layer 3) switch, and the like. It is also called an external network that is different from a network including switches) through the switches.
As servers are virtualized and comprehensively managed in configuration of such a cloud system, efficiency of operation is sometimes improved by virtually unifying plural switches into one and unifying control and management of plural routes.
Here, a switch side provides networks used by each VM (Virtual Machine) as communication environment of layer 2, because VMs that operate on servers in virtual environment are ordinarily interconnected by layer 2. Moreover, it is general to provide layer 2 communication environment among VMs by using VLAN (Virtual Local Area Network) or the like, because the networks used by each VM (a broadcast range of layer 2) are different from each other.
In present server virtualizing systems, there are several methods that define a VLAN between a server and a port of a switch that is directly connected to the server, and a VLAN definition in a boundary point between the server and the port is automatically or manually set according to a state of a VM that is used.
On the other hand, when servers communicate with external networks through a core network, a setting of the same VLAN as a VLAN that has been defined for ports of switches that are directly connected to the servers is applied to ports of switches that are connected to the core network. However, there is no proper method for setting a VLAN like this automatically.
Patent Document 1: Japanese Laid-open Patent Publication No. 09-284328
Patent Document 2: Japanese Laid-open Patent Publication No. 09-284329